That is unexpected.
If $OSTYPE is an environment variable, then it should be accessible in G’MIC (it does a getenv() for that, in l.4781 of gmic.cpp).
So I suspect your environment variable is not passed to gmic, which is a bit insane, as that should be the role of an environment variable (so that executables can detect in which environment they run).
Another Mac weirdness ?
At this point, I have no other ways to detect that gmic is running on darwin!
What about a user defined one e.g. GMIC_MAC ? I hesitated already if this strong solution is good for all other mac users. Those with one large monitor might prefer the comfortable placement and size solution for this mode?
Maybe a boolean GMIC_MAC or some sort of GMIC_OSTYPE=$OSTYPE could help. Still I hesitate to have to system specific things. The latter would be quite near the linux system!
Maybe we can use the result of uname to detect a MacOSX system inside G’MIC, so we can avoid having to define an additional variable.
What the output of uname for you ?
#@cli is_macosx
#@cli : Return 1 if current computer OS is Darwin (MacOSX), 0 otherwise.
is_macosx :
if !narg($_is_macosx) l[]
file=${-path_tmp}gmic_uname
x "uname >"$file
it $file
_is_macosx={same(crop(),'Darwin',6,0)}
rm
onfail _is_macosx=0 rm
endl fi
u $_is_macosx
Ok, result is “1” under zsh and bash, but that should not matter. Hopefully nobody will try “dp” under the plugin, app environment is quite different compared to terminal processes!
That is strange. Might be something to do with the new test done by the is_macos command.
I’ll just made an attempt to fix this, will be ready in a few minutes.